Despite the development of new medications to treat alcohol dependence (AD), pharmacotherapy remains an underutilized approach to alcohol treatment. This is, in large measure, a function of modest effect sizes associated with the FDA-approved medications for AD. The present proposal is to extend findings from a randomized double-blind clinical trial of ondansetron, a specific serotonin-3 (5-HT3) antagonist, in European- American (EA) and Hispanic individuals with AD. In that trial, the medication was efficacious in reducing drinking among individuals homozygous for the L allele of the 52-HTTLPR in the serotonin transporter gene (SLC6A4), an effect that was enhanced in individuals who were T homozygotes for the 32-UTR SNP (rs1042173) in the same gene. The proposed study will address important limitations to that study by balancing individuals with the 32-UTR SNP across the LL genotype, making it possible to characterize accurately the degree to which ondansetron's therapeutic effect was enhanced by the TT genotype and to determine the independent therapeutic effect of ondansetron in individuals with the TT genotype compared with their genotypic counterparts (i.e., TG and GG). In addition, we will do the proposed study with an equal number of EAs and African-Americans (AAs) to enhance the generalizability of the findings to the general population. Finally, we will use the tri-allelic genotype (L2 vs. S2 allele) for the 52-HTTLPR to avoid underestimation of the pharmacogenetic effect associated with the use of the bi-allelic 52-HTTLPR genotype (L vs. S alleles). Our overarching hypothesis is that ondansetron, by modulating the effects of polymorphic variation at SLC6A4 through postsynaptic 5-HT3 receptor down-regulation, will decrease the frequency of heavy drinking in individuals homozygous for the L2 and T alleles at the 52-HTTLPR and the 32-UTR SNP, respectively. This effect will be seen in two distinct populations: EAs and AAs. That is, individuals who are L2L2/TT will exhibit the most robust therapeutic response. To test this hypothesis, we will enroll only treatment-seeking, AD individuals with the L2L2 genotype, who will be assigned to 2 sets of 4 cells (total N = 8), each containing 32 individuals (total N=256). Because of the need to recruit a very large sample (N=~1,000) from which to randomize only L2 homozygotes, the study will be conducted at both the University of Virginia and the University of Pennsylvania. Subjects will be allocated into one of 2 groups-AAs and EAs. Each group, containing only L2 homozygotes, will be randomized in a 2 (TT vs. TG or GG) x 2 (ondansetron 4 mg/kg twice daily vs. placebo) factorial design in a 16-week clinical trial. Subjects also will receive weekly brief behavioral compliance enhancement treatment (BBCET) as their psychosocial adjunct. The successful demonstration of genetic moderation of the effects of ondansetron, which can be expected to yield a moderate effect size, will help to personalize the treatment of AD and demonstrate the great potential benefit in the pharmacological treatment of the disorder. PUBLIC HEALTH RELEVANCE: This proposal is to extend findings from a randomized double-blind clinical trial of ondansetron, in which the medication was found to reduce drinking among individuals with certain genotypes (i.e., forms of DNA, the material that controls the inheritance of characteristics). The proposed study will address a number of limitations in the prior work, including testing the medication in both European-American and African-American samples. To ensure an adequate number of participants, the study will be conducted at both the University of Virginia and the University of Pennsylvania, by researchers with considerable experience in the conduct of such studies.